The aim of this paper is to design an IoTs (Internet of Things) system that can detect failures on distribution transformers based on the temperature of their body. To build the system, at first, a device is deployed to measure the temperature outside the transformer using a non-contact infrared temperature sensor. Then, experiment is conducted to collect the temperature data of operating transformers to find out a correlation between the outside temperature and actual temperature of them to conduct suitable warning mechanism. Next, Amazon AWS IoT Core service is utilized to connect measuring devices and store their data in a database. Finally, a web-based application is developed to monitor and alert the abnormal states of transformers. The experimental results indicate that the system can measure transformer temperature based on the heat on their body, the web-based application can observe transformer temperature in real-time and notify when a problem is occured.